An open-source honkong-based reverse car searching method, device and medium

Abstract

The application discloses a reverse vehicle searching method and device based on open source Hongmeng, and a medium. The method comprises the following steps: acquiring relative measurement data between beacons arranged on the top of each parking space based on the star flash protocol of open source Hongmeng, and dynamically networking the beacons according to the relative measurement data to construct a two-dimensional relative position lattice; identifying a preliminary screening lane area according to the spatial interference relationship between virtual rectangles formed by bidirectional extension along the normal direction of the collinear cluster; identifying a real lane area meeting the traffic condition in the preliminary screening lane area, generating a hybrid topology road network diagram according to virtual nodes and beacons in the real lane area for representing a traffic track; in response to a vehicle searching request for a target vehicle, generating a corresponding vehicle searching track according to the location of a target beacon node; and issuing a light control instruction to the corresponding beacon node to guide the user to the target vehicle through the output light of the beacon node.

Description

Technical Field

[0001] This application relates to the field of indoor navigation technology, specifically to a reverse vehicle finding method, device, and medium based on the open-source HarmonyOS. Background Technology

[0002] With the surge in motor vehicle ownership, the scale of underground parking lots in large urban commercial complexes and transportation hubs is constantly expanding, making it difficult to find one's car a core pain point. However, existing car-finding systems rely on pre-constructed high-precision electronic maps of parking lots to achieve accurate navigation. Due to the complex environment of underground parking garages, the construction of such maps requires a significant amount of manual on-site surveying or importing computer-aided design (CAD) files, resulting in long construction cycles and high costs. Furthermore, when parking spaces are re-marked or passageways are temporarily blocked due to construction in the parking garage, the electronic map cannot automatically detect environmental changes and must be manually revised, leading to extremely high maintenance costs. In addition, most existing car-finding systems render navigation routes on electronic maps on mobile phone screens, making the system heavily reliant on the user's mobile network. Large underground parking garages often have severe signal blind spots, and the car-finding system will be unusable if the network is interrupted or congested. Summary of the Invention

[0003] To address the aforementioned issues, this application proposes a reverse vehicle search method based on the open-source HarmonyOS, comprising: Based on the StarFlash protocol of open source HarmonyOS, relative measurement data between beacons deployed on the top of each parking space is obtained, and the beacons are dynamically networked according to the relative measurement data to construct a two-dimensional relative position matrix. Collinear clusters are extracted from the two-dimensional relative position lattice. For two adjacent collinear clusters, the spatial interference relationship between the virtual rectangles formed by bidirectional extension along the normal direction of the collinear clusters is used to identify the initial screening lane area. Identify the actual lane areas that meet the traffic conditions in the initially screened lane areas, and generate a hybrid topology road network map based on the virtual nodes used to represent the traffic trajectory in the actual lane areas and the beacons; The target vehicle is bound to the target beacon node in the hybrid topology road network map. In response to a vehicle search request for the target vehicle, a corresponding vehicle search trajectory is generated based on the location of the target beacon node. Based on the user's real-time location as they move along the vehicle-finding trajectory and the type of the trajectory segment to which the real-time location belongs, a light control command is sent to the corresponding beacon node to guide the user in finding their vehicle through the output lights of the beacon node.

[0004] In one implementation of this application, the beacons are dynamically networked based on the relative measurement data to construct a two-dimensional relative position matrix, specifically including: Select the first beacon from the beacons as the first management node; Using the first management node as the access starting point, other terminal nodes corresponding to beacons are accessed sequentially. During the access process, cross constraints are applied to the terminal nodes to be accessed based on the relative measurement data between the already accessed nodes and the terminal nodes to be accessed. Repeat the above access process until all nodes corresponding to all beacons have been connected, and obtain a global list of discovered nodes with completed network formation; Determine the initial two-dimensional coordinates corresponding to each node in the global discovered node list, and perform global nonlinear least squares optimization on the initial two-dimensional coordinates to obtain the globally optimized two-dimensional relative position lattice. A geometric morphology scan is performed on the globally optimized two-dimensional relative position lattice to identify collinear clusters. Based on the virtual rigid straight line constraint terms constructed from the collinear clusters, the globally optimized two-dimensional relative position lattice is iteratively aligned to obtain the optimized two-dimensional relative position lattice.

[0005] In one implementation of this application, the relative measurement data includes distance and azimuth. Based on the relative measurement data between the connected nodes and the terminal node to be connected, cross constraints are applied to the terminal node to be connected, specifically including: The first terminal node that can directly establish a Starlink link with the first management node; When the first management node continues to connect to the second terminal node, a reference circle is constructed with the first management node and the first terminal node as the center and the distance between the center and the second terminal node as the radius. Based on the intersection of the reference circle and the azimuth angle of the second terminal node relative to the first terminal node, the coordinates of the second terminal node are determined and the second terminal node is connected. After connecting to the second terminal node, for the terminal node to be connected, at least two connected nodes with direct ranging links to the terminal node to be connected are selected, and geometric cross constraint relationships are constructed based on the relative measurement data between the connected nodes and the terminal node to be connected. Based on the aforementioned geometric intersection constraints, the terminal nodes to be connected are sequentially accessed.

[0006] In one implementation of this application, based on the geometric intersection constraint relationship, the terminal nodes to be connected are sequentially accessed, specifically including: Based on the geometric intersection constraint relationship, the candidate coordinates corresponding to the terminal node to be connected are solved; Calculate the spatial deviation value between each group of candidate coordinates. If the spatial deviation value is less than a preset threshold, then perform a weighted average on each group of candidate coordinates, and connect the terminal node to be connected based on the coordinate value obtained after weighted averaging.

[0007] In one implementation of this application, collinear clusters are extracted from the two-dimensional relative position lattice. For two adjacent collinear clusters, the initial screening lane regions are identified based on the spatial interference relationship between the virtual rectangles formed by bidirectional extension along the normal direction of the collinear clusters. Specifically, this includes: Linear fitting is performed on the two-dimensional relative position lattice, and the set of collinear beacons with a spacing equal to the preset parking space width is extracted as the collinear cluster of the same row of parking spaces; For each collinear cluster, starting from each beacon in the collinear cluster, virtual rectangles with two directions of positive and negative extension are generated along the normal direction according to the preset parking space size. The spatial interference relationship between virtual rectangles of adjacent collinear clusters under different orientations is detected, and the true orientation of each collinear cluster is identified based on the spatial interference relationship; wherein, the spatial interference relationship includes the interval distance and the overlap state; Based on the actual orientation, the continuous blank areas between adjacent collinear clusters are identified as the initial screening lane areas.

[0008] In one implementation of this application, identifying the actual lane areas that meet the traffic conditions in the initially screened lane areas specifically includes: Extract the centerline of the initially screened lane area as the initial geometric skeleton line; Calculate the total distance between each node set at a fixed interval on the initial geometric skeleton line and the boundary of the virtual rectangle located on both sides of it, and use the total distance as the local passage gap; Cut off skeleton line segments in the initial geometric skeleton line where the local passage gap is less than a preset safe passage threshold. For the remaining geometric skeleton line after cutting off the skeleton line segments, perform connectivity detection on the remaining geometric skeleton line to use the connected skeleton as the real lane area.

[0009] In one implementation of this application, a hybrid topology road network map is generated based on the virtual nodes used to represent travel trajectories in the real lane area and the beacons, specifically including: At fixed intervals, several virtual nodes are generated along the centerline of the actual lane area; Each beacon is orthogonally projected onto the adjacent real lane area, and a topological connection is established between the beacon and the virtual node at its projection point. Adjacent virtual nodes are then connected in series to generate a hybrid topological road network map composed of the real nodes and virtual nodes corresponding to the beacons.

[0010] In one implementation of this application, a light control command is sent to the corresponding beacon node based on the user's real-time location while moving along the vehicle-finding trajectory and the type of the trajectory segment to which the real-time location belongs. Specifically, this includes: When the trajectory segment type is a straight lane segment, a light control command is issued to the beacon nodes in the collinear clusters on both sides of the straight lane segment to control the beacon nodes to display the corresponding guide color; When the trajectory segment type is a turning segment, a directional light command is issued to the beacon node in front of the user's field of vision to control the beacon node to display a directional light array. The device is characterized by comprising: At least one processor; And, a memory communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, which are executed by the at least one processor to enable the at least one processor to execute a reverse vehicle search method based on open-source HarmonyOS as described above.

[0011] This application provides a non-volatile computer storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured as follows: One of the above is a reverse vehicle search method based on the open-source HarmonyOS.

[0012] The reverse vehicle search method based on open-source HarmonyOS proposed in this application can bring the following beneficial effects: A two-dimensional relative position matrix is ​​constructed using star-flash ranging between beacons. This matrix automatically identifies lane and parking space areas and generates a hybrid topology road network map. This eliminates the reliance on manually mapped electronic maps found in traditional solutions, achieving zero mapping costs for deployment. Furthermore, when the parking lot environment changes, the beacon network can automatically sense and update the topology through re-ranging, without manual intervention. During the vehicle location guidance phase, the planned path is mapped to dynamic changes in the lights of beacon entities. The light signals guide the user to the target parking space, eliminating the need for the user to check their phone screen or rely on any network signal, thus improving driving safety and vehicle location convenience. Attached Figure Description

[0013] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart illustrating a reverse vehicle search method based on open-source HarmonyOS provided in this application embodiment; Figure 2 A schematic diagram of a virtual rectangle provided for an embodiment of this application; Figure 3 This is a schematic diagram of a reverse vehicle finding device based on the open-source HarmonyOS, provided as an embodiment of this application. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0015] The following, in conjunction with the accompanying drawings, details one reverse vehicle search method based on the open-source HarmonyOS in each embodiment of this application, including: S101: Based on the open-source HarmonyOS StarFlash protocol, it acquires relative measurement data between beacons deployed on the top of each parking space, and dynamically networks the beacons according to the relative measurement data to construct a two-dimensional relative position matrix.

[0016] To enable reverse parking by autonomously constructing a spatial topology within a parking lot without the need for a pre-installed electronic map, beacons need to be installed at the top center of each parking space's entry side, i.e., the guide side for vehicles entering the parking space, typically directly above the intersection of the parking space line and the lane line. The beacon installation height should be consistent. Under this arrangement constraint, the physical location of the beacons throughout the parking lot presents a regular array of grid dots, with the spacing between beacons in adjacent parking spaces within the same row approximately equal to the width of a standard parking space. In this embodiment, the beacon is an integrated device equipped with the open-source HarmonyOS operating system and a built-in high-precision ranging module, possessing ranging, angle measurement, and communication capabilities. After power-on deployment, the beacons, based on the open-source HarmonyOS soft bus and the StarScan protocol, can autonomously generate a two-dimensional relative position dot matrix map in a blind state without a pre-installed map or global coordinate system, relying solely on the relative distance measurements between the beacons deployed above each parking space. This is achieved through dynamic networking and coordinate deduction, providing a spatial topology foundation for subsequent lane recognition and navigation.

[0017] In one embodiment, after the beacons are powered on, all beacons automatically initiate networking based on the open-source HarmonyOS distributed soft bus and the StarScan protocol. The system selects a beacon located in a corner and conforming to the deployment rules as the first management node. Using this as the origin of the relative coordinate system, all other beacons are considered as terminal nodes T. The first management node serves as the access starting point, broadcasting a discovery message via the StarSignal protocol to sequentially access the terminal nodes T corresponding to other beacons. It should be noted that as terminal nodes connect, the first management node will monitor the air interface signal-to-noise ratio of the edge terminal nodes in real time. When a remote terminal node... Due to physical obstruction (by load-bearing columns, ventilation ducts) and When the signal-to-noise ratio of the starburst signal falls below a preset threshold, the terminal node is automatically promoted to a new management node, denoted as [node name missing]. It is responsible for taking over the discovery of other nodes. Each management node synchronizes the global list of discovered nodes in real time through the open-source HarmonyOS distributed soft bus until the total number of nodes in the list equals the preset total number of beacons installed in the system. At this point, the underlying radio frequency network seamlessly closes the coverage.

[0018] During the process of successively connecting terminal nodes, this embodiment of the application no longer uses the single-link cascading derivation method to expand nodes, i.e., A connects to B, B connects to C. Instead, it adopts multi-point cross constraints. Based on the relative measurement data between the connected nodes and the terminal nodes to be connected, cross constraints are applied to the terminal nodes to be connected to ensure that the local topology is not distorted. The relative measurement data includes measured distance and azimuth angle.

[0019] Specifically, the first management node uses star-based ranging and angle measurement to determine its first neighbor, i.e., the first terminal node, with which it can directly establish a star-based link. With the origin as the point, With the first directly connectable terminal node The measured distance between them is When connecting a second terminal node, multipath reflections and occlusions in the underground parking garage may lead to inaccurate single-point measurements. Therefore, it is necessary to introduce the already connected first terminal node as a reference for cross-validation and correction. First, determine the first management node. With the second terminal node Measured distance between The measured distance between the first terminal node and the second terminal node Using the first management node and the first terminal node as centers, two reference circles are constructed with the measured distance from each center to the second terminal node as the radius. With the center of the circle, Draw a circle with radius 1. ,by With the center of the circle, Draw a circle with radius 1. The intersection of the two reference circles generates two candidate intersection points. By combining the azimuth information of the second terminal node relative to the line connecting the first management node and the first terminal node, the coordinates of the second terminal node can be uniquely determined from the two candidate intersection points. At this point, the access of the second terminal node can be completed.

[0020] Furthermore, as the network continues to expand outward and requires access to subsequent terminal nodes further away, the system selects at least two already connected nodes from the currently connected nodes that have direct ranging links with the terminal node to be connected. Using the relative measurement data between each already connected node and the terminal node to be connected, multiple sets of independent geometric cross-constraint relationships are constructed. In this embodiment, the geometric cross-constraint relationship can be a triangular constraint relationship. Based on the geometric cross-constraint relationship, each set of constraints independently calculates the candidate coordinates corresponding to the terminal node to be connected. Based on the candidate coordinates, the final coordinates of the terminal node can be further selected, thereby allowing it to connect to the current network. Specifically, assuming that there already exists a... , , The reference triangle formed is for the terminal node to be connected. Select , , Forming a triangle Using relative measurement data , and known edges Solve using the Law of Cosines Relative to the edge - The first set of candidate coordinates Then, select , , Forming a triangle Using relative measurement data , and known edges Solution Relative to the edge - The second set of candidate coordinates After obtaining the two sets of candidate coordinates, the spatial deviation between each set of candidate coordinates is calculated using the following formula: If the spatial deviation is less than a preset threshold (typically 0.1m), a weighted average is calculated for each group of candidate coordinates. The weight is inversely proportional to the shortest side of the triangle. Based on the coordinates obtained after the weighted average, the location and access of the terminal node to be connected can be completed. Conversely, if the spatial deviation is greater than or equal to the preset threshold, it is determined that there is abnormal ranging data affected by multipath interference. In this case, the ranging data corresponding to the geometric intersection constraint relationship that causes the largest deviation needs to be removed; that is, the data corresponding to the geometric intersection constraint relationship that causes the largest deviation needs to be removed. and Zhongyu For triangle constraints with significant deviations, the solution is re-executed using other connected nodes. If no other available nodes are found, the connection of that terminal node is temporarily suspended.

[0021] In subsequent node expansion, the above access process is repeated continuously. Each node synchronizes the globally discovered node list in real time through the open-source HarmonyOS distributed soft bus, ensuring the consistency and globality of the network topology perception among management nodes. The above process is repeated level by level according to the wavefront expansion order from near to far, until all terminal nodes are assigned initial coordinates, resulting in a globally discovered node list for network formation. Compared with the traditional single-link cascade derivation method, the above multi-point cross-constraint access mechanism calls multiple already connected nodes for redundancy verification in the positioning of each new node. Multipath deviations of a single ranging edge are promptly detected and suppressed by other constraints, thereby ensuring the accuracy of the local topology structure and completely avoiding topology collapse caused by single-point errors.

[0022] After obtaining the global list of discovered nodes, each node is assigned initial two-dimensional coordinates. However, since all nodes are calculated based on relative coordinates, the accumulation of small measurement errors during the long-term, step-by-step expansion process may cause slight rotation, stretching, or twisting deformation of the distal endpoint array. To eliminate this global deformation, after all nodes have obtained their initial two-dimensional coordinates, these coordinates need to be globally optimized.

[0023] Specifically, all beacons are treated as graph vertices, and the measured distances between beacons are considered graph edges. Global nonlinear least squares optimization is performed starting from the initial two-dimensional coordinates. This involves finding a set of global coordinates that minimizes the sum of squared errors between the calculated and measured distances on all edges. During this optimization process, edges with higher ranging signal-to-noise ratios and more reliable measurements are assigned higher weights, while edges severely affected by multipath interference have lower weights. This effectively suppresses the negative impact of abnormal ranging on the overall optimization, resulting in a more accurate overall geometry of the point matrix.

[0024] To further eliminate cumulative deformation at large scales, this application introduces a topological feature self-identification and rigid alignment mechanism. After the first round of global distance optimization converges, no preset physical environment is made; instead, a purely mathematical geometric morphological scan is performed on the generated two-dimensional lattice. Since the beacons are densely deployed in a grid-like pattern in physical space, a large number of high-confidence collinear node clusters will inevitably appear in the lattice—that is, several beacons that are geometrically approximately located on the same straight line, corresponding to the set of beacons at the top of the same row of parking spaces in physical space. Through purely geometric means such as calculating the eigenvalues ​​of the covariance matrix of the node set, these collinear features are automatically identified and abstracted into virtual rigid straight-line constraints, which are then added to the optimization model for secondary iterative alignment. This virtual rigid straight-line constraint is not an actual rigid object or measured edge in physical space, but rather a geometric pattern autonomously identified by the system from the lattice morphology and transformed into a mathematical constraint. In other words, these nodes, which should be strictly collinear, are applied to the optimization model. In this secondary optimization, collinear nodes that were originally slightly bent or skewed due to accumulated errors are pulled to a uniform straight-line direction by this constraint, completely correcting any possible torsional deformation of the lattice as a whole.

[0025] The above global optimization process is specifically expressed by the following formula:

[0026] Where E represents the set of all links in the system that can establish star-flash ranging; This represents a single ranging link between nodes i and j, and p represents the iteratively optimized planar coordinates. This indicates the actual physical distance measured by the ranging module. , Representing an edge The estimated value of the ranging variance is inversely derived from the HADM ranging signal-to-noise ratio (SNRij) of the ranging module, i.e. The more reliable the ranging, the greater the weight; conversely, the weight of edges severely affected by multipath interference is reduced, thus suppressing their impact on the overall optimization. This represents the weighting coefficient, used to balance the ranging error loss and the topological collinearity alignment loss. K represents the total number of high-confidence collinear node clusters automatically identified by the system through geometric scanning. , , The standard equation of the kth virtual rigid line x+ y+ The coefficient is equal to 0. , This represents the optimized coordinates.

[0027] Through the aforementioned dynamic networking, multi-point intersection constraint access, global least squares optimization, and collinear rigid alignment, a high-fidelity and distortion-free two-dimensional relative position matrix map can be output solely through relative measurements and mathematical deductions between beacons, without the need for any manual surveying or pre-set maps. This provides a spatial topology foundation for the subsequent automatic identification of lane areas and road network generation.

[0028] S102: Extract collinear clusters from the two-dimensional relative position lattice. For two adjacent collinear clusters, identify the initial screening lane area based on the spatial interference relationship between the virtual rectangles formed by bidirectional extension along the normal direction of the collinear clusters.

[0029] The two-dimensional relative position point matrix output by the above steps consists of discrete coordinate points. It's impossible to determine which nodes belong to the same row of parking spaces, the orientation of the parking spaces, or the location of the lanes. Therefore, it's necessary to analyze the geometric distribution characteristics of the two-dimensional relative position point matrix to identify passable lanes. By performing linear fitting on the two-dimensional relative position point matrix, collinear clusters are extracted, and the normal direction of these collinear clusters is determined as the entry extension direction for that row of parking spaces. Since this normal direction can have both positive and negative orientations, the system draws virtual rectangles along both directions according to the preset parking space size. By detecting the spatial interference relationship between virtual rectangles of adjacent collinear clusters under different orientations, the initial screening of lane areas is identified.

[0030] In one embodiment, a linear fitting is performed on the two-dimensional relative position matrix. Since the beacons are physically installed precisely at the top center of each parking space's entry side, the beacon spacing between adjacent spaces in the same row is approximately equal to the standard parking space width. Therefore, the beacons in the same row naturally exhibit an approximately collinear geometric distribution in the matrix. The system performs linear fitting based on the spatial distribution relationship between the coordinate points, identifying the set of beacons that are geometrically approximately collinear and whose adjacent spacing equals the preset parking space width as a collinear cluster within the same row of parking spaces. This collinear cluster corresponds to all the beacons installed on the top of a row of parking spaces in the physical space. The fitted line direction of this collinear cluster is calculated, and the direction perpendicular to this fitted line direction is taken as the normal direction. This normal direction represents the entry extension direction of that row of parking spaces, i.e., the direction of travel when a vehicle enters that row of parking spaces.

[0031] However, relying solely on the fitted straight lines of the collinear cluster can only determine the spatial linear position of the beacon arrangement. Whether the orientation of the parking space is positive or negative along the normal direction presents two completely opposite possibilities at a purely geometric level. To resolve these two orientation assumptions, virtual rectangles extending in both positive and negative directions are generated along the normal direction according to preset parking space lengths and widths, starting from the coordinates of each beacon in the collinear cluster. These virtual rectangles provide a precise geometric simulation of the physical parking space's location on a two-dimensional plane, with each virtual rectangle's length and width corresponding to the actual physical dimensions of the parking space.

[0032] After generating virtual rectangles, the spatial interference relationship between virtual rectangles of adjacent collinear clusters under different orientation combinations is detected. This spatial interference relationship includes the spacing and overlap between rectangles. The true orientation of each row of beacons is determined by analyzing the spatial positional relationship between the virtual rectangles of adjacent collinear clusters. Figure 2 As shown in the schematic diagram of a virtual rectangle, the length of the virtual rectangle is L and the width is W. If two rows of rectangles do not overlap in a certain direction and the width of the continuous blank area exposed in the middle is greater than the lane width threshold, such as... and If the rectangular area is filled, the blank area is identified as the initial screening lane area, and the true orientation of each row of beacons under that orientation is determined. Conversely, if a row of beacons, when drawing a virtual rectangle using a certain orientation assumption, collides spatially with a lane area already locked by the above interference detection or another virtual rectangle with a confirmed orientation, it proves that the orientation assumption violates the physical space constraints, and the system forcibly flips the orientation of that row of beacons to the opposite direction. For example, when the distance between two adjacent collinear clusters is approximately equal to twice the length of a parking space, if the width of the blank area exposed between the rectangles after the two rows adopt the opposite pointing assumption is 2L, which far exceeds the maximum limit width of the physical lane, then the opposite pointing assumption is overturned. At this time, the system switches the two rows to the opposite extension assumption. If the two rows of rectangles have a slight overlap with an area less than a threshold or a tiny gap with a width less than the width of a single parking space, the system determines that the spatial feature is a tail-to-tail double-row parking space abutment structure and locks the true orientation of the two rows of beacons to the opposite direction. When the two rows of beacons adopt the assumption of pointing towards each other, the two rows of virtual rectangles do not overlap, and the continuous blank area in the middle with a width greater than W is identified as the initial screening lane area.

[0033] S103: Identify the real lane areas in the initially screened lane areas that meet the traffic conditions, and generate a hybrid topology road network map based on the virtual nodes and beacons used to represent the traffic trajectory in the real lane areas.

[0034] Virtual rectangles divide the plane containing the two-dimensional relative position lattice into a physical occupant area and a preliminary screening lane area. However, due to mapping errors or incomplete alignment of parking spaces, some pseudo-connected narrow gaps may exist in the preliminary screening lane area, with insufficient physical width to accommodate vehicle passage. These gaps appear connected but are actually too narrow for vehicles to pass through. Therefore, it is necessary to further identify the actual lane areas that truly meet the passage conditions within the preliminary screening lane area, and generate a hybrid topology road network map based on the virtual nodes and beacons used to represent the passage trajectory in the actual lane areas.

[0035] In one embodiment, the centerline of the initially screened lane blank area is extracted as the initial geometric skeleton line, which is a geometric representation of the lane center travel trajectory. Several nodes are set at fixed intervals on the initial geometric skeleton line. Each node of the initial geometric skeleton line is traversed, and the sum of the vertical distances from each node to the nearest virtual rectangle boundaries on both sides is calculated. This sum of distances is defined as the local clearance of the node, used to quantify the width that vehicles can pass through at that location. A preset safe passage threshold is set, typically the maximum vehicle width plus a safety margin. If the local clearance at a certain location is less than the preset safe passage threshold, it means that the width at that location is insufficient for vehicles to pass safely, and the skeleton line segment at that location needs to be cut. After cutting, a connectivity test is performed on the remaining geometric skeleton lines. If a skeleton segment becomes an isolated line segment not connected to other skeleton segments due to cutting, it is determined to be a pseudo lane generated by error and is removed. Conversely, if the skeleton remains connected after cutting, it proves that although the area is locally narrowed, it is still a passable real lane. The finally retained connected skeleton is the confirmed real lane area. This completes the precise verification of the lane area, eliminating impassable narrow gaps caused by errors and confirming a truly continuous passage space for vehicles.

[0036] In one embodiment, after obtaining the actual lane area, several virtual nodes are automatically generated at fixed intervals along the central axis of the area. Each virtual node is assigned a unique logical identifier and precise two-dimensional coordinates in the digital twin topology to represent the discrete travel position on the lane. Based on this, and using the prior constraint that the beacon is located at the center point of the parking space entry side, the position of each physical beacon is orthogonally projected onto the adjacent actual lane area, and a topological connection is established between the beacon and the virtual node at its projection point to form the access trajectory between the parking space and the lane. At the same time, adjacent virtual nodes are connected in series to form the straight-line trajectory edge of the lane, and the virtual nodes at the intersection of the skeleton lines are interconnected in a mesh to form the turning trajectory edge of the lane. Through this fusion, a hybrid topological road network map is generated, which is composed of physical beacon nodes and virtual road network nodes. The hybrid topological road network map has both the physical anchoring capability of beacon entities and the topological connectivity expression capability of virtual nodes, providing a complete spatial topological structure for subsequent vehicle binding and route planning.

[0037] S104: Bind the target vehicle to the target beacon node in the hybrid topology road network map, and in response to the vehicle search request for the target vehicle, generate the corresponding vehicle search trajectory based on the location of the target beacon node.

[0038] After the hybrid topology road network map is constructed, in order to realize reverse vehicle search, it is necessary to bind the user's target vehicle with the corresponding target beacon node in the hybrid topology road network map, and plan a vehicle search trajectory from the current location to the target vehicle for the user when a vehicle search request is received.

[0039] This application provides a dual-channel vehicle binding mechanism to adapt to user habits and deployment needs under different parking lot management conditions. First, users can actively bind their vehicles on the terminal side. After parking, users input or confirm their vehicle's license plate number using an atomic service card on a mobile terminal running the open-source HarmonyOS system. Then, through star-flash ranging interaction between the terminal and surrounding beacons, the nearest beacon node to the user's terminal is determined as the vehicle's parking location, and a unique association is established between the license plate information and the physical beacon node. Simultaneously, the binding relationship is uploaded to cloud storage. Second, binding can be performed via cloud-based vision. For parking lots with deployed license plate recognition cameras, the camera equipment above the barrier gate or parking space automatically recognizes the license plate number and the parking space number when a vehicle enters and parks. The cloud system directly and automatically binds the license plate to the corresponding beacon node above the parking space, without any manual operation from the user. These two channels can be flexibly selected or coexisted depending on the actual hardware configuration of the parking lot. Regardless of the method used, a one-to-one mapping relationship can be established between the identity information of the target vehicle and the beacon node above the physical parking space where it is parked, so that the physical beacon node in the road network map is given a specific vehicle identity semantic.

[0040] When a user returns to the parking lot and needs to find their vehicle, they initiate a vehicle search request using an open-source HarmonyOS atomic service card. Upon receiving the search command containing the target license plate information, the system first retrieves the target beacon node bound to the license plate in the hybrid topology road network map based on established binding relationships. This node represents the precise location of the target vehicle in the road network space. Through star-flash ranging interaction between the user terminal and surrounding beacons, the system measures the distance between the user's current location and multiple surrounding beacons. The system uses the beacon node closest to the user in the hybrid road network map as the starting point for pathfinding, and the target beacon node as the endpoint. It then runs a shortest path algorithm on the hybrid topology road network map, which is composed of interwoven virtual and physical nodes, to calculate the optimal vehicle search trajectory consisting of a continuous sequence of topological nodes.

[0041] S105: Based on the user's real-time location while moving along the vehicle search trajectory and the type of the trajectory segment to which the real-time location belongs, send a light control command to the corresponding beacon node to guide the user in finding the vehicle through the output lights of the beacon node.

[0042] Once a user initiates a vehicle search request and completes trajectory planning, the system needs to guide the user from their current location along the planned trajectory to the target vehicle's location. Existing navigation methods mostly achieve this by rendering an electronic map and displaying the navigation path on the phone screen. Users must constantly look down at the screen, which poses safety hazards and is prone to frequent failures due to network signal blind spots in underground parking garages. Therefore, this application's embodiment transforms navigation commands into dynamic changes in the lighting of beacon entities, using the user's perception of changes in surrounding environmental lighting during walking to achieve screenless wayfinding guidance.

[0043] As the user moves along the vehicle search trajectory, the user terminal continuously interacts with surrounding beacons to measure distances using starlight. Based on this, the user's current position on the road network map is tracked in real time, and the type of trajectory segment to which the position belongs is determined. The trajectory segment type includes straight lane segment, turning segment, or end approach segment. Different types of trajectory segments correspond to different lighting control strategies.

[0044] Specifically, when a user is in a straight lane, the system extracts the nearest collinear beacon clusters on both sides of the lane and issues a light control command to simultaneously illuminate the beacons on both sides with a unified guide color, thus constructing a light corridor in the physical space. The user can move along the light corridor based solely on visual instinct, without needing to look down at any screen. Simultaneously, based on the user's real-time position and progress along the trajectory, the light corridor is controlled to sequentially roll forward and illuminate, then gradually dim backward, creating a continuous sense of dynamic progression. If, during the user's movement, it is detected that the user has deviated from the centerline of the light corridor and moved closer to one side, the beacons on that side are immediately switched to a warning color distinct from the guide color, visually alerting the user to correct their direction of travel, achieving real-time deviation correction in the physical space.

[0045] When a user is about to enter a turning section, relevant beacons within a certain range in front of the user's field of vision are activated in advance, and directional light commands are issued to control the beacons to display light arrays with clear directional indications, such as directional arrows or flowing light patterns, to guide the user to turn naturally at the turning point instead of going straight, ensuring that the user receives the turning prompt before reaching the turning point and avoids missing the intersection.

[0046] When the topological hop count between the user and the target beacon node in the road network map is less than or equal to the preset hop count threshold, it means that the user has approached the area where the target vehicle is located. At this time, the system sends an end-of-line disclosure command to the target beacon node, controlling the beacon to switch to a specific color that is different from the guide color and remain constantly lit or to perform high-frequency flashing. By marking the parking space where the target vehicle is located, the system helps the user quickly locate the vehicle in the final range.

[0047] By employing the aforementioned differentiated dynamic lighting guidance strategy that divides the route into segments and stages, the continuous guidance trajectory is discretized into a series of lighting control commands that can be executed by the beacon hardware. This enables visual navigation throughout the entire physical space from the starting position to the target parking space without requiring the user to view their mobile phone screen or access a mobile network signal.

[0048] The above are embodiments of the methods proposed in this application. Based on the same idea, some embodiments of this application also provide devices corresponding to the above methods.

[0049] Figure 3 This is a schematic diagram of the structure of a reverse vehicle finding device based on the open-source HarmonyOS provided in an embodiment of this application. Figure 3 As shown, it includes: At least one processor; and, At least one processor-communication-connected memory; wherein, The memory stores instructions that can be executed by at least one processor, which enables the at least one processor to perform the reverse vehicle search method based on open-source HarmonyOS as described in any of the preceding items.

[0050] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device and medium embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the description of the method embodiments.

[0051] The devices and media provided in this application are one-to-one with the methods. Therefore, the devices and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0052] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0053] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0054] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0055] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0056] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0057] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0058] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0059] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0060] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A reverse vehicle search method based on open-source HarmonyOS, characterized in that, The method includes: Based on the StarFlash protocol of open source HarmonyOS, relative measurement data between beacons deployed on the top of each parking space is obtained, and the beacons are dynamically networked according to the relative measurement data to construct a two-dimensional relative position matrix. Collinear clusters are extracted from the two-dimensional relative position lattice. For two adjacent collinear clusters, the spatial interference relationship between the virtual rectangles formed by bidirectional extension along the normal direction of the collinear clusters is used to identify the initial screening lane area. Identify the actual lane areas that meet the traffic conditions in the initially screened lane areas, and generate a hybrid topology road network map based on the virtual nodes used to represent the traffic trajectory in the actual lane areas and the beacons; The target vehicle is bound to the target beacon node in the hybrid topology road network map. In response to a vehicle search request for the target vehicle, a corresponding vehicle search trajectory is generated based on the location of the target beacon node. Based on the user's real-time location as they move along the vehicle-finding trajectory and the type of the trajectory segment to which the real-time location belongs, a light control command is sent to the corresponding beacon node to guide the user in finding their vehicle through the output lights of the beacon node.

2. The reverse vehicle search method based on open-source HarmonyOS according to claim 1, characterized in that, Based on the relative measurement data, the beacons are dynamically networked to construct a two-dimensional relative position matrix, specifically including: Select the first beacon from the beacons as the first management node; Using the first management node as the access starting point, other terminal nodes corresponding to beacons are accessed sequentially. During the access process, cross constraints are applied to the terminal nodes to be accessed based on the relative measurement data between the already accessed nodes and the terminal nodes to be accessed. Repeat the above access process until all nodes corresponding to all beacons have been connected, and obtain a global list of discovered nodes with completed network formation; Determine the initial two-dimensional coordinates corresponding to each node in the global discovered node list, and perform global nonlinear least squares optimization on the initial two-dimensional coordinates to obtain the globally optimized two-dimensional relative position lattice. A geometric morphology scan is performed on the globally optimized two-dimensional relative position lattice to identify collinear clusters. Based on the virtual rigid straight line constraint terms constructed from the collinear clusters, the globally optimized two-dimensional relative position lattice is iteratively aligned to obtain the optimized two-dimensional relative position lattice.

3. The reverse vehicle search method based on open-source HarmonyOS according to claim 2, characterized in that, The relative measurement data includes distance and azimuth. Based on the relative measurement data between the connected nodes and the terminal nodes to be connected, cross constraints are applied to the terminal nodes to be connected, specifically including: The first terminal node that can directly establish a Starlink link with the first management node; When the first management node continues to connect to the second terminal node, a reference circle is constructed with the first management node and the first terminal node as the center and the distance between the center and the second terminal node as the radius. Based on the intersection of the reference circle and the azimuth angle of the second terminal node relative to the first terminal node, the coordinates of the second terminal node are determined and the second terminal node is connected. After connecting to the second terminal node, for the terminal node to be connected, at least two connected nodes with direct ranging links to the terminal node to be connected are selected, and geometric cross constraint relationships are constructed based on the relative measurement data between the connected nodes and the terminal node to be connected. Based on the aforementioned geometric intersection constraints, the terminal nodes to be connected are sequentially accessed.

4. The reverse vehicle search method based on open-source HarmonyOS according to claim 3, characterized in that, Based on the aforementioned geometric intersection constraints, the terminal nodes to be connected are sequentially accessed, specifically including: Based on the geometric intersection constraint relationship, the candidate coordinates corresponding to the terminal node to be connected are solved; Calculate the spatial deviation value between each group of candidate coordinates. If the spatial deviation value is less than a preset threshold, then perform a weighted average on each group of candidate coordinates, and connect the terminal node to be connected based on the coordinate value obtained after weighted averaging.

5. The reverse vehicle search method based on open-source HarmonyOS according to claim 1, characterized in that, Extract collinear clusters from the two-dimensional relative position lattice. For two adjacent collinear clusters, identify the initial lane regions based on the spatial interference relationship between the virtual rectangles formed by extending bidirectionally along the normal direction of the collinear clusters. Specifically, this includes: Linear fitting is performed on the two-dimensional relative position lattice, and the set of collinear beacons with a spacing equal to the preset parking space width is extracted as the collinear cluster of the same row of parking spaces; For each collinear cluster, starting from each beacon in the collinear cluster, virtual rectangles with two directions of positive and negative extension are generated along the normal direction according to the preset parking space size. The spatial interference relationship between virtual rectangles of adjacent collinear clusters under different orientations is detected, and the true orientation of each collinear cluster is identified based on the spatial interference relationship; wherein, the spatial interference relationship includes the interval distance and the overlap state; Based on the actual orientation, the continuous blank areas between adjacent collinear clusters are identified as the initial screening lane areas.

6. The reverse vehicle search method based on open-source HarmonyOS according to claim 1, characterized in that, Identifying the actual lane areas that meet the traffic conditions within the initially screened lane area specifically includes: Extract the centerline of the initially screened lane area as the initial geometric skeleton line; Calculate the total distance between each node set at a fixed interval on the initial geometric skeleton line and the boundary of the virtual rectangle located on both sides of it, and use the total distance as the local passage gap; Cut off skeleton line segments in the initial geometric skeleton line where the local passage gap is less than a preset safe passage threshold. For the remaining geometric skeleton line after cutting off the skeleton line segments, perform connectivity detection on the remaining geometric skeleton line to use the connected skeleton as the real lane area.

7. The reverse vehicle search method based on open-source HarmonyOS according to claim 1, characterized in that, Based on the virtual nodes used to represent travel trajectories in the real lane area and the beacons, a hybrid topology road network map is generated, specifically including: At fixed intervals, several virtual nodes are generated along the centerline of the actual lane area; Each beacon is orthogonally projected onto the adjacent real lane area, and a topological connection is established between the beacon and the virtual node at its projection point. Adjacent virtual nodes are then connected in series to generate a hybrid topological road network map composed of the real nodes and virtual nodes corresponding to the beacons.

8. The reverse vehicle search method based on open-source HarmonyOS according to claim 1, characterized in that, Based on the user's real-time location while moving along the vehicle-finding trajectory and the type of the trajectory segment to which the real-time location belongs, a light control command is sent to the corresponding beacon node, specifically including: When the trajectory segment type is a straight lane segment, a light control command is issued to the beacon nodes in the collinear clusters on both sides of the straight lane segment to control the beacon nodes to display the corresponding guide color; When the trajectory segment type is a turning segment, a directional light command is issued to the beacon node in front of the user's field of vision to control the beacon node to display a directional light array.

9. A reverse vehicle location device based on the open-source HarmonyOS, characterized in that, The device includes: At least one processor; And, a memory communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform a reverse vehicle search method based on open-source HarmonyOS as described in any one of claims 1-8.

10. A non-volatile computer storage medium storing computer-executable instructions, characterized in that, The computer-executable instructions are set as follows: A reverse vehicle search method based on open-source HarmonyOS as described in any one of claims 1-8.

Images